Journal of Molecular Structure 744–747 (2005) 621–626 www.elsevier.com/locate/molstruc
Spectroscopic studies of glassy phospho-silicate materials Maciej Sitarza,*, Mirosław Handkea, Zbigniew Fojudb, Stefan Jurgab a
AGH University of Science and Technology, Faculty of Materials Science and Ceramics Al. Mickiewicza 30, 30-059 Krako´w, Poland b Institute of Physics, Adam Mickiewicz University, Umultowska 65, 61-614 Poznan´, Poland Received 30 November 2004; accepted 29 December 2004 Available online 17 February 2005
Abstract The aim of the work is to determine the internal structure of bioactive glasses based on structural studies. Due to the absence of the longrange order, the X-ray methods usually applied in the studies of crystalline materials are of low applicability in the investigations of glasses. Therefore, spectroscopic methods, such as IR and NMR, which make it possible to ‘see’ the short- and the middle-range order are extremely suitable in their studies. IR investigations have shown that the glasses studied exhibit non-uniform domain composition, which should facilitate their transformation into glass-crystalline state. Additionally, detailed NMR measurements have been carried out in order to determine the influence of glass structure on the course of its crystallization. These studies make it possible to verify the common opinions on the middle-range order (domains) in the glasses studied. The results obtained for 27Al and 31P are particularly promising. Changes in the environment of [SiO4]4Ktetrahedra, which can be spatially connected with various numbers of [PO4]3K and [AlO4]5K tetrahedra cause the presence of various types of bridging bonds, such as Si–O–Si, Si–O–Al and Si–O–P. This in turn has significant impact on the number and the shape of the bands occurring in the spectra. q 2005 Elsevier B.V. All rights reserved. Keywords: Glass ceramics; MIR spectra; NMR spectra; phospho-silicate materials
1. Introduction The glassy-crystalline materials obtained via direct crystallization from the glassy state are one of the most interesting ceramic materials. Material of the predetermined size distribution of crystalline grains and the crystallization degree should be the product of directed crystallization. At present, directed crystallization of glass is considered as one of the most promising methods of preparation of nanomaterials, which can be applied as materials for photonics, magnetic and ferromagnetic materials, catalytic supports as well as bioactive materials for medicine [1–3]. Application of glass as the nanomaterial precursor allows to take the advantages of specific properties of glassy state, especially the ease to control properties via the appropriate selection of the chemical composition and the possibility to use various preparation methods. Design of new glasses and glass-ceramic * Corresponding author. Tel.: C48 12 172232; fax: C48 12 331593. E-mail address:
[email protected] (M. Sitarz). 0022-2860/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2004.12.050
materials requires the detailed analysis of the glass structure. Amorphous silico-phosphate glasses of NaCaPO4–SiO2 and NaCaPO4–AlPO4–SiO2 systems have been studied. Glassy and glassy-crystalline materials of this system with such additives, as e.g. KC, MgC2, SrC2, BC3 are the basis for synthesis of various bioactive, i.e. capable of forming bonds with tissue, materials. Microscopic as well as EDX studies of glasses of NaCaPO4–SiO2 and NaCaPO4–AlPO4– SiO2 systems have shown that liquation exists in the materials obtained. A significant and characteristic effect of aluminium on composition of both, the matrix and inclusions has been noticed and described. For all materials of NaCaPO4–SiO2 system spherical inclusions are practically pure silico-phosphate phase, whereas the matrix is enriched in sodium, calcium and phosphorus. In the case of materials belonging to NaCaPO4–SiO2–AlPO4 system spherical inclusions contain exclusively the sodium–calcium–phosphate–phosphate phase, whereas the matrix is a pure silicate phase for the samples of the highest SiO2 content. When the amount of NaCaPO4 exceeds 25 mol%
622
M. Sitarz et al. / Journal of Molecular Structure 744–747 (2005) 621–626
Table 1 Composition of the samples (mol%) Series I
NaCaPO4–SiO2 system
Series II
AlPO4–NaCaPO4–SiO2 system
I Ia Ib Ic Id Ie
100% SiO2 10% NaCaPO4, 90% SiO2 20% NaCaPO4, 80% SiO2 30% NaCaPO4, 70% SiO2 40% NaCaPO4, 60% SiO2 50% NaCaPO4, 50% SiO2
II IIa IIb IIc IId IIe
100% SiO2 5% AlPO4, 5% NaCaPO4 90% SiO2 5% AlPO4, 15% NaCaPO4, 80% SiO2 5% AlPO4, 25% NaCaPO4, 70% SiO2 5% AlPO4, 35% NaCaPO4, 60% SiO2 5% AlPO4, 45% NaCaPO4, 50% SiO2
chemical composition of the inclusions changes and they become silicate phase. Composition of the matrix also alters but in the opposite way [4,5]. It should be noted that liquation is an advantageous phenomenon in view of the preparation of glassy-crystalline materials since spherical inclusions play the role of crystal nuclei.
2. Experimental Glasses belonging to NaCaPO4–SiO2 and NaCaPO4– AlPO4–SiO2 systems whose compositions are given in Table 1 have been selected for the studies. Bioactivity and existence of liquation in these systems have been the main criteria for the choice of these compositions [6–8]. The sol–gel method was selected to obtain the materials of the highest possible homogeneity. TEOS (SiO2), Ca(NO 3) 2.4H2 O (CaO), Na 3PO 4.12H 2O (Na 2 O) and H3PO4 (P2O5) were used to introduce particular oxides. The obtained gels were dried at room temperature (30 days) and then at the temperature of 60 8C. After drying process all the gels were heated at 1380 8C to obtain crystalline materials. The attempts of glassy samples obtaining on the way of progressive heating of gels were unsuccessful. Finally the gels were melting in platinum crucible in the temperature of 1700 8C and rapidly cooled on the cast iron plate. X-ray measurement of all samples were carried out using FPM Seifert XRD 7 with a step of 0,01 deg and collecting time-5 s. EDX spectra were measured on a JEOL 5400 scanning microscope with microprobe analyser LINK ISIS (Oxford Instrument). IR spectroscopic measurements of the resulting materials were made with a Bio-Rad FTS 60 V spectrometer. Spectra were collected after 256 scans at 4 cmK1 resolution. Samples were prepared by the standard KBr pellets method. The, 27Al, and 31P NMR chemical shifts are reported in ppm scale from adequate reference lines. All NMR spectra were recorded using a Bruker DSX 400 MHz spectrometer operating at 104.3, 161.9 MHz for 27Al and 31P NMR, respectively. Samples were held in 4 mm zirconia rotors and spun at 15 kHz.
3. Results and discussion Knowledge of the structure of glass constituting the precursor of glassy-crystalline material is the basis for the proper design of directed crystallization process. Therefore, as precise as possible determination of the structure of glasses obtained has been the main goal of the present work. Amorphous nature of the materials prepared has been established using X-ray diffraction studies. Characteristic raised background in the range about 20–35K2Q visible in the XRD patterns of the materials containing the highest amount of SiO2 (Ia–Ic and IIa–IIc) proves their amorphous nature. In the case of materials containing the highest amount of NaCaPO4 (Id–Ie and IId–IIe) distinct maxima on the raised background appear, which show that there exists a small amount of crystalline phase in the systems.
Fig. 1. FIR spectra of samples Ia–Ie.
M. Sitarz et al. / Journal of Molecular Structure 744–747 (2005) 621–626
623
range ordering (domains) close to that occurring in the corresponding crystalline materials. Microscopic and EDX investigations have shown that liquation exists in the glasses studies—spherical inclusions in the matrix [4,5]. Since MIR spectra give the averaged information on the sample it is not possible to establish liquation in the studied glasses. Based on combined MIR, EDX and XRD studies it is possible, however, to conclude indirectly that depending on composition ordering within inclusions is similar either to crystalline phosphate phase or crystalline crystobalite. In the case of Ia–Ic samples inclusions are silicate phase, whereas the matrix is enriched in sodium, calcium and phosphorus [4,5]. Thus, based on MIR spectra, it can be indirectly concluded that the short range ordering within the inclusions is similar to that of crystalline crystobalite, whereas within the matrix it resembles that of crystalline NaCaPO4. Similar conclusions concerning short range ordering within the domains of the matrix and inclusions can be drawn in the case of glassy materials belonging to NaCaPO4–AlPO4–SiO2 system–samples IIa–IIc. In the case of MIR spectra corresponding to amorphous materials precise determination of basic band parameters (position, intensity, full width at half maximum) is possible only after decomposition of the broad bands occurring in the spectra into their components. In the studies of glassy Fig. 2. FIR spectra of samples IIa–IIe.
These findings have been confirmed by spectroscopic studies in far infrared region (FIR). FIR spectra of the obtained materials are presented in Fig. 1 and 2. No bands are visible in the spectra of the Ia–Ic and IIa–IIc samples which unequivocaly shows their amorphous nature. In the case of Id–Ie and IId–IIe samples one can distinguish a number of bands which proves the presence of crystalline phases. Only fully amorphous materials have been selected for further studies (samples Ia–Ic and IIa–IIc). This selection has been based on the results of XRD and FIR investigations. Owing to very broad bands the spectra of glassy materials in the middle infrared region (MIR) are difficult to interpret. Therefore, for their interpretation the spectra of the corresponding crystalline materials are necessary. MIR spectroscopic investigations, which have been carried out for crystalline materials of selected systems has made it possible to identify the bands originating from Si–O and P–O bonds vibrations [9]. MIR spectra of the amorphous materials selected for further studies together with their crystalline analogues are presented in Figs. 3 and 4. In all the spectra corresponding to amorphous materials the bands exhibit very high full widths at half maximum, which again proves their amorphous nature. High similarity of the spectra measured for amorphous and crystalline materials shows that the amorphous materials studied exhibit short
Fig. 3. MIR spectra of crystalline and amorphous samples of the NaCaPO4SiO2 (Ia–Ic) system.
624
M. Sitarz et al. / Journal of Molecular Structure 744–747 (2005) 621–626
Fig. 5. Decomposition of MIR spectrum of Ib sample.
Fig. 4. MIR spectra of crystalline and amorphous samples of the NaCaPO4– SiO2-AlPO4 (IIa–IIc) system.
silico-phosphate materials decomposition of the MIR spectra is necessary also because the bands due to [SiO4]4K and [PO4]3K tetrahedra overlap. Identical procedure, proposed in the work [10] for all spectra decomposition was applied. In Figs. 5 and 6 decomposition of two selected spectra (samples Ib and IIb) are presented. The most intensive and broad band in the spectra of all amorphous samples at ca. 1300–900 cmK1 is a combination of Si–O and P–O stretching vibrations. The shape of this band can be modified by the presence of double SiaO bonds, nonbridging Si–OK bonds and double PaO bonds [11]. Decomposition of the spectra into components allows to assign the bands at 1144 and 1168 cmK1 in the spectra of Ib and IIb samples to the SiaO and PaO stretching vibrations. The bands at 1109 and 1102 cmK1 as well as those at 1042 and 1047 cmK1 (Figs. 5 and 6) are connected with the bridging stretching vibrations of Si–O(P) and Si–O(Si). The presence of the band at 935 and 938 cmK1 in both decomposed spectra should be also noted. This band implies depolymerization of silico–oxygen network resulting from the presence of modifiers–NaC and CaC2 ions. This band can unequivocally be assigned to the stretching Si–OK (NaC, CaC2) vibrations. The influence of aluminium on the structure of the studied glasses is manifested in the spectrum corresponding to sample IIb (decomposition—Fig. 6) by
lower number of the bands. Significant differences in the intensities of the bands (intensity expressed as the peak area) are also visible. Aluminium affects particularly the intensity of the bands connected with the SiaO type network defects (the bands at 1144 cmK1—sample Ib and 1168 cmK1—sample IIb) as well Si–OK (NaC, CaC2) (the bands at 935 and 938 cmK1). Significant lowering of these bands in the spectrum of sample IIb (i.e. the one containing aluminium) indicates that the occurrence of non-tetrahedral cations in the vicinity of aluminium is favoured. Owing to
Fig. 6. Decomposition of MIR spectrum of IIb sample.
M. Sitarz et al. / Journal of Molecular Structure 744–747 (2005) 621–626
625
Fig. 8. 31P MAS NMR spectra of Ia–Ic samples.
Fig. 7. 27Al MAS NMR spectra of IIa–IIc samples.
similarities in the nature of Si–O and Al–O bonds as well as close values of Al and Si atomic masses it is not possible to resolve the bands due to Al–O (aluminium in tetrahedral coordination) and Si–O vibrations in the MIR spectra even after their decomposition into component peaks. The bands characteristic for the vibrations of Al–O bonds (aluminium in octahedral coordination) are not observed in the MIR spectra of the studied glasses, either. 27Al MAS NMR investigations, which have been carried out (Fig. 7) have shown that aluminium is present exclusively in tetrahedral coordination—the intensive band at C54 ppm [12]. NMR studies make it possible to establish unequivovally that aluminium forms the glass network. Peak asymmetry, particularly pronounced in the spectrum of sample IIb (Fig. 7), indicates that aluminium occurs in two nonequivalent positions. This may be related to the liquation of the studied glasses. Microscopic studies combined with EDX have shown that aluminium occurs mostly in inclusions, which are mainly sodium–calcium-phosphate phase. It has been found, however, that small amounts of aluminium are present also in silicate matrix [5]. Appearance of two different non-tetrahedral cations (NaC and Ca2C) in the vicinity of aluminium may also be the reason for result in the asymmetric peak in the NMR spectrum. Incorporation of aluminium in tetrahedral coordination ([AlO4]5K tetrahedra) into silicate structure leads to the excessive negative charge. This explains why non-tetrahedral cations in the surrounding of aluminium are preferred. The conclusion that nontetrahedral cations in the vicinity of alumino-oxygen tetrahedra are favoured can be also drawn from the MIR spectra (Figs. 5 and 6)— marked lowering in the intensity of the band at ca. 935 cmK1 due to Si–OK (NaC, Ca2C) vibrations. The effect of aluminium on the structure of the materials obtained can be also observed in the 31P MAS NMR spectra— Figs. 8 and 9. It is particularly distinct in the 31P MAS NMR spectra: in those corresponding to Ia–Ic samples
(not containing aluminium) practically one peak at ca. C 4 ppm is seen. Based on the literature data [13,14] it can be ascribed to P(Q1) species, where Q—number of oxygen bridging atoms. In the spectra corresponding to Ib and Ic samples significant asymmetry of the peak at ca.K4 ppm is observed. This may indicate the presence of P(Q2) species [13,14]. Addition of aluminium at low NaCaPO4 content (sample IIa, Fig. 9) leads to almost exclusive presence of P(Q2) species: the main peak at K5 ppm. A distinct shoulder at ca. C4 ppm implies the occurrence of a small amount of P(Q1) species. As the content of NaCaPO4 increases (samples IIb and IIc, Fig. 9) the intensity of the peak at ca. K5 ppm sharply decreases (in the case of the spectrum corresponding to IIc sample this peak completely disappears) with simultaneous–similarly sharp–increase in the intensity of the band at ca.K4 ppm. Results of 31P MAS NMR studies correlate well with EDX investigations, which have shown the inversion of the composition of the matrix and inclusions in the case of IIc sample [5]. When the content of NaCaPO4 exceeds 25( the sample containing aluminium (IIc) exhibits practically the same structure as that without aluminium (Ic).
Fig. 9.
31
P MAS NMR spectra of IIa–IIc samples.
626
M. Sitarz et al. / Journal of Molecular Structure 744–747 (2005) 621–626
4. Conclusions
Acknowledgements
1. MIR spectroscopic investigations of crystalline and amorphous silico-phosphate materials together with decomposition of the spectra into component bands, which have been carried out have made the detailed assignments of the bands to the appropriate types of vibrations possible. 2. It has been shown that the short range ordering in the studied silico-phosphate amorphous materials is similar to that characteristic for the corresponding crystalline materials. It has been also found that already a small addition of aluminium (5% mol AlPO4) to the amorphous silico-phosphate materials significantly influences their structure by lowering the number of SiaO and Si–OK (NaC, CaC2) defects. 27 3. Al MAS NMR studies have shown that in the investigated materials aluminium is present exclusively in tetrahedral coordination and it takes part in the creation of glass network. It has been also established (together with MIR studies) that in the samples containing aluminium non-tetrahedral cations occur in the vicinity of alumino-oxygen tetrahedra. 4. 31P MAS NMR investigations have made it possible to confirm the significant influence of aluminium on the structure of the investigated materials. In the materials which do not contain aluminium almost exclusively P(Q 1) species occur, whereas in those to which aluminium is added (at low content of NaCaPO4) P(Q2) species prevail.
This work is supported by Polish Committee for Scientific Research under grant no. 3T08D 033 27 Maciej Sitarz is a scholarship holder of The Foundation For Polish Science (Scholar Grant 2003 and 2004).
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
H.G. Kim, T. Komatsu, J. Mat. Sci. Lett. 17 (1998) 1198. T. Niskida, S. Kubuki, J. Mater. Chem. 7 (1997) 1801. L.R. Puckney, G.H. Beal, J. Non-Cryst Solids. 219 (1997) 219. M. Sitarz, M. Rokita, M. Handke, E. Galuskin, J. Mol. Struct. 489 (2003) 651–653. M. Handke, M. Sitarz, M. Rokita, E. Galuskin, J. Mol. Struct. 39 (2003) 651–653. L.L. Hench, R.J. Splinter, W.C. Allen,T, K. Greenlee, J. Biomed. Mater. Res. 5 (1972) 117. L.L. Hench, H.A. Paschall, J. Biomed. Mater. Res. Symp. 4 (1973) 25. G. Piotrowski, L.L. Hench, W.C. Allen, J. Biomed. Mater. Res. 9 (1975) 47. M. Sitarz, M. Rokita, M. Handke, J. Mol. Struct. 614 (2002) 289. M. Handke, W. Mozgawa, M. Nocun´, J. Mol. Struct. 555 (2000) 299. M. Rokita, W. Mozgawa, M. Handke, J. Mol. Struct. 596 (2001) 171. G.L. Turner, R.J. Kirkpatrick, S.H. Risbud, E. Oldfield, Am. Ceram. Soc. Bull. 66 (1987) 656. H. Yamashita, K. Nagata, H. Yoshino, K. Ono, T. Maekawa, J. NonCryst Solids 248 (1999) 115. H. Yamashita, H. Yoshino, K. Nagata, H. Inoue, T. Nakajin, T. Maekawa, J. Non-Cryst Solids 248 (1999) 115.
